A STUDY O F T H E HYDROGEN ELECTRODE O F THE CALOMEL ELECTRODE AND O F CONTACT POTENTIAL. V BY W. F. CLARKE, C. N. MYERS AND S. F. ACREE
I. INTRODUCTION The increasingly accurate results which have been obtained by various workers who have studied the constancy, reproducibility, and, especially, the applicability of the hydrogen electrode and of the calomel electrode, made the continued study of these systems seem well worth while. With the idea of continuing the work on this subject that was begun in Acree’s laboratory by Desha,’ and continued by Loomist2and Myers3 with very interesting results, this investigation was undertaken. The work of Hildebrand4 in his study of the application of the hydrogen electrode as an instrument for determining the titration curves for various acids, bases and salts, that of Schmidt and Finger5 in their study of the “Potential of a Hydrogen Electrode in Acid and Alkaline Solutions,” and that of Bjerrum6 in connection with his work on the “Elimination of Contact Potentials” show a few of the important phases of this work along inorganic lines. The hydrogen electrode was used by Denham’ in measuring the degree of hydrolysis of several inorganic salts, and of aniline hydrochloride. The chief object of the study of the hydrogen electrode in this laboratory has been to devise a method for allowingus to investigate a number of such catalytic reactions as the following : Desha: Diss., Johns Hopkins Univ., 1909; Am. Chem. Jour., 46, 638 (1911). a
Loomis: Ibid., 1911;.4m. Chem. Jour., 46, 585-622 (1911). Am. Chem. Jour., 50, 396 (1913). Jour. Am. Chem. SOC.,35, 847-871 (1913). Jour. Phys. Chem., 12, 406-416 (1908). Zeit. Elektrochemie, 17, 389-393 (1911). Jour. Chem. SOC., gj, 41 (1908).
244
W . F. Clarke, C. N . M j ’ e r s a d S. F . Acree
CH3CONHz
++ HC1
+
CHsCOSHs
+ HzO
-
+ C1
+
+
-
CHaCOOH I& C1 It is at once apparent that there are so many unknowns in this reaction that the conductivity and reaction velocity methods together cannot allow us to determine whether the cation or nonionized salt or both are hydrolyzed, but if we could measure a t any moment the concentration of the hydrogen ions by means of the hydrogen electrode, we could solve this problem. Conductivity measurements by Nirdlinger showed that there can be a t most only a few percent of the amide salt and it is therefore necessary to develop the hydrogen electrode to a high degree of reproducibility, if i t is to be of service in this connection. As an illustration we can choose a solution in which five percent of the hydrogen ions combine with the amide to form the amide salts. In order to show a change of 0.1percent in the hydrogen ion concentration and hence z percent in the concentration of the amide salt, we must make the measurements accurate to 0.000025 volt. h-ow it is highly desirable in such studies to know the concentration of the amide salts t o within 0.1percent, and hence we must be able to measure e. m. f. with an accuracy of O.OOOOOI volt. The smaller the percent of hydrogen ions disappearing in such salt-formation, the greater the accuracy demanded in the e. m. f . measurement. We have tried to devise methods to allow us to approach a reproducibility of O.OOOOOI volt. In studying the hydrogen electrode, the calomel electrode has been used as a standard; consequently the latter needed to be studied with respect to its constancy and ease of reproduction. This involves five lines of study, which, briefly, are: I. The determination of the average value of the potential of a large number of electrodes, arranged in batteries of ten, which were made up very carefully from chemicals purified by the best known methods, while the apparatus was kept as clean as practicable. IT. The best methods for preparing and keeping the hydrogen electrodes in condition for immediate use. CH3COXH3C1
A S t u d y of the H y d r o g e n Electrode, Etc.
245
111. Intercomparisons of different electrodes and systems. IV. Modifications of the apparatus, especially with reference t o the heating system, the bath, new forms of apparatus for calomel electrodes, and new hydrogen electrode apparatus. V. Measurement of the systems (a)Hs-Pt-o. I A' HC1-o. I S KCI-HgC1-Hg. ( b ) 0.1 LYKCl-HgC1-Hg. (c) 0 .I -\-HC1-HgC1-Hg. ( d ) Hg-HgC1-0. I HC1-0. I KC1-HgC1-Hg. ( e ) Hs-Pt-0.1 S HCl. ( j ) H2-Pt-0. I S HC1-0. I HC1-HgC1-Hg. (g) Hg-HgC1-o. I HC1 sand filling 0 . I KC1-HgC1-Hg. A\T
11. HISTORICAL
Loomis gave a brief summary of the previous work in the study of calomel and hydrogen electrodes in which he stated that Ostwald? was the first t o describe the normal calomel electrode, which he found reproducible to within one millivolt and to have the potential +o.j600 0.0006 i t " IS") as determined by the drop-electrode method by Rothm ~ n d .The ~ study of Coggeshal14on the constancy of calomel electrodes ; that of Srna1e.jon gas elements, in which he showed that the electromotive force of the cell is independent of the material in the electrode, if not acted upon chemically; and that of JVilsmore, Richards, Gewecke,s Sauer, and others
+
These newest forms are not closed by constricting and sealing the glass a t the top as were those designed by Myers and Acree (Am. Chem. Jour., 5 0 , 398, footnote), nor are the ground-glass tops used by Loomis employed. The top of each single electrode has a ground-glass stopper, and a glass cover tube is placed over this and fits down on a small bulb blown in the upper tube of the electrode, t o which bulb the cover tube is attached hermetically with sealing wax. A full description appears later in this article. Ostwald-Luther : "Physiko-chemische blessungen," 3rd Edition, p. 441. Zeit. phys. Chem., 15, 15 (1894). Ibid., 17,62 (1895). Ibid., 14, 577 (1894). 6 Ibid., 35, 296 (1900); 40, 385 (1902). 'I Ibid., 24, 39 (1897). * Ibid., 45, 685 (1903). Ibid., 47, 146 (1904).
W . F . Clarke, C. N . Myers and S . F . Acree
246
are discussed briefly by Loomis, besides which he gives an extensive bibliography. Wilsmore2 tabulated the results of the studies of Smale, Xeumann, and himself. He found for the potential of the normal calomel electrode the mean value 0.283 volt which was based on his adoption as a zero of potential that of the hydrogen electrode towards a solution normal with respect to hydrogen ions. The elements studied by these workers and from which this result was calculated, included the following systems : ( a ) H2-Pt-1 N "21-1 N HC1-HgC1-Hg ( b ) H2-Pt-1 N HCI-I N KC1-HgCl-Hg (c) H2-Pt-1 N HCI-0.5 N KCI-0.5 N KC1-HgC1-Hg ( d ) H2-Pt-I N H2S04-I N KC1-HgCl-Hg (e) H2-Pt-1 N H2S04-0.5 N KC1-HgC1-Hg Especial mention should be made in this brief historical survey, of the work of S ~ h o c h S, ~a ~ e rSalm,j ,~ Planck,6 Palmaer,' Nernst, * Lorenz and Mohn, Lewis and Sergeant,lo Abegg and Cumrnings,l' and especially that of Bjerrum.12
111. THEORETICAL The potential of the hydrogen electrode toward the solution in which it is immersed is dependent upon the pressure of the hydrogen gas and upon the osmotic pressure of the hydrogen ions in the solution, as was shown by Kernst. The theory regarding the calculation of the electromotive force between a calomel electrode and a hydrogen electrode is explained by Loomis, as follows: Loomis: p. j j , Diss., Johns Hopkins University, 1911. Zeit. phys. Chem., 35, 296 (1900). 8 Jour. Phys. Chem., 14,665 (1910). 4 Zeit. phys. Chem., 47, 146 (1904). 6 Zeit. Elektrochemie, IO, 341 (1904). 8 Wied. Ann., 40, j 6 1 (1890). 7 Zeit. phys. Chem., 59, 129 (1909). 8 Ibid., 56, 544 (1906). 9 Ibid., 60,,422 (1907). 10 Jour. Am. Chem. SOC.,31,363 (1909). 11 Zeit. Elektrochemie, 13, 17 (1907). Ibid., 17,58 (1911). 1 2
A Study o j the H3'drogen Electrode, Etc.
247
"In the comparison of a calomel electrode against a hydrogen electrode in a solution whose hydrogen ion concentration is H', we find that if 7r represents the observed electromotive force, 7r1 the potential of the calomel electrode against a hydrogen electrode when immersed in a solution with unit concentration of hydrogen ions, F the quantity of electricity necessary to deposit one gram ion of hydrogen and 7r2 the contact potential between the solutions of the system, then the equation
x
RT
=
7rl
-F
log, H'
+
r2
holds when the hy-
drogen gas is under atmospheric pressure. 273 ") we find that when T = (25 O
+
"The value of the potentiometer, T~
and
RT
x 7r1
From this equation
is obtained by actual measurement with is calculated from some system in which
loglo H' are known and
7r
has been previously
measured, and 7r2 is calculated. The best system for determining the value of 7rl is: H2-Pt 0.1lY HCI, 0.11 Y KCl HgClHg in which
T
can be measured,
RT
log, H' = 0.0591 X
loglo 0.0922 and xa can be calculated by some such formula as that of Planck." The calculation of 1r2 is difficult and methods have been sought for eliminating this contact potential between the two systems. For this purpose there has been interposed between the two solutions in question a saturated solution of a very soluble salt, whose anions and cations have nearly equal migration velocities. Abegg and Cummingl used for this purpose ammonium nitrate. Denham assumed they were correct in claiming that the use of ammonium nitrate practically eliminates contact potential. He used it in measuring the hydrolysis of aniline hydrochloride. Desha and Loornis showed that it does not entirely eliminate contact potential. The sand-filling method of Bjerrum for eliminating con1
Zeit. Elektrochemie, 13, 17 (1907).
248
W . F . Clarke, C. N . Myers and S. F . Acree
tact potential is much more efficient, as shown by Myers. The apparatus required for it consists of a Peligot tube filled with very pure sand to within an inch and a half of the top of each branch. The one side is tightly stoppered and 0.1-qi KCl is poured in very cautiously and allowed to soak through thoroughly until it reaches the center line of the small bulb a t the bottom of the tube. A siphon carrying a stopcock and filled with 0.1,"L' KC1 is now fitted tightly by means of a ground joint into the side of the Peligot tube, which has been soaked with the potassium chloride solution. The stopper is removed from the other side and 0. I h' HCl is introduced with care, so that the hydrochloric acid and potassium chloride solutions come in contact in a sharp line in the middle of the lowest bulb of the Peligot tube. A siphon of 0.1S HC1 is attached in the manner described above to the hydrochloric acid side. When a hydrogen electrode battery is to be connected to a 0.1 KC1 calomel battery for intercomparison, the siphon of potassium chloride is introduced into the calomel battery and the siphon of hydrochloric acid is connected with the hydrogen electrode battery. In case the calomel battery is made up with 0.1 HC1 for an electrolyte and sand-filling device is not necessary ; the hydrogen battery can be connected directly to the calomel battery, as illustrated at J, in Fig. I . -!*
*!-
IV. EXPERIMENTAL 1. Apparatus The use of the potentiometer for measuring electromotive force; of the Weston cells as primary standards; of an oil bath for maintaining constant temperature, the oil eliminating electric leakage currents far better than water; and of the electrodes themselves, is in the main the same as that described by Loomis with the exception of such modifications and additions as are described below. There are now in use five Weston standard cells loaned by the Bureau of Standards. The use of the potentiometer and Weston cells as arranged a t present permits an accuracy of within o.oo001 volt.
A Stud31
o j the Hydrogen Electrode, Etc.
249
The temperature regulation is correct to within 0.01 O, the experiments being carried out at 25'. An especially designed milk-scale thermometer is used. It is made of Jena 59 I11 glass and consists of a 2 cc bulb and a capillary tube having small bulbs between the fixed points o", 5 O , IO', 15', zoo, 2 5 O , 30" and 3 5 O . On each side of the fixed points the milk scale is graduated in 0.02' over a range of 0.2". By the use of an especially arranged cathetometer we can easily read the thermometer to 0.0005°. The bulb wall was made rather thicker than usual in order to diminish the errors usually incident to Beckmann thermometers having thin walls. By having the bulbs in the capillary the length 2 inches. This type has of the thermometer is reduced to 1 6 ~ proved to be very reliable, the observed temperatures being reproducible to within 0.003' at 25' and 0.005' a t 35". In place of the electric bulb used by Loomis for heating there has been introduced a heating coil consisting of nichrome wire wound in a spiral upon insulated posts, and divided by lead wires into three sections. The heating capacity of each section is equivalent to that of a 32 candle power light. These sections may be thrown in series or parallel by means of a switch board. It is found that better regzclatioiz o j tentperatwe and electromotive force is obtained by their use for the reason that the coils are exposed directly to the oil and retain the heat for a shorter time than does a lamp. The thermo-regulation is accomplished by means of the toluene regulator and the master relay system, first devised in Acree's laboratory for maintaining a given temperature constant to within a few thousandths of a degree. In the master-relay system, the toluene thermo-regulator and a tmovolt storage circuit actuate a 250 ohm Bunnell relay, or a very sensithe No. 18 n'eston relay, which, in turn, operates sex-era1 other relays acting as short-circuits for the current for the heating coil. Each spark gap is short-circuited by a condenser and a high resistance lamp to prevent sparking or arcing. The stirring of the oil is accomplished by a fan stirrer
250
W . F . Clarke, C.iV. Myers and S. F . Acree
which drives the oil down over the heating coil whence i t passes under a glass plate supported four inches above the bottom of the bath; the oil flows to the further end of the bath and returns after passing with good contact through the gridiron toluene thermo-regulator, above the glass plate which supports the apparatus; it is during the last stage of the trip that it comes in contact thoroughly with the apparatus. During the passage of the oil under the glass plate the oil passes over pipes through which cold water is circulated when the external temperature renders it desirable. By means of this system the regulation was constant to within 0.01'. The stirrer ic) circulating the oil is attached to the shaft A in Fig. I , which also carries, above the surface of the oil,
Fig.
I
the fan ( b ) for stirring the air. A closely fitting top of wood and glass covers the entire bath. The shaft carries the friction wheel ( d ) which, with the shaft and fans, is revolved rapidly by means of the small friction wheel (e) attached to the shaft of the motor. The motor (f) is supported just outside the end of the bath by an iron frame work (g) which is bolted, through holes in supporting brass plates, t o the table on which the bath rests. This arrangement avoids vibrations great enough to interfere appreciably with reading the galvanometer
A S / u d y o j the Hydrogeyz Electrode, E t c .
251
on the wall at the farther end of the room. T h i s system avoids the troubles incideizt to the apparatus used b y L o o m i s , such as the breakittg a n d slipping 08o j belts, bad tibration and the open bath. The galvanometer is supported on a marble slab which is cemented in the wall so as to avoid as far as possible the effect of jarring caused by machinery elsewhere in the building. The cover consists of a frame work of glass and wood which is held in position over the bath by strips of brass which project for several inches along the outside of the bath at the four corners. It has a top which is as long as the bath and is hinged to the main frame of the cover about four inches from the top back edge. This top consists of two parts hinged together in such a way that it will readily fold back upon itself and permit free access to the interior of the bath. At the end of the bath a t which the motor for stirring is located, there are arranged two sliding doors which have had such sections cut from them as will permit the passage of the shaft of the motor. The doors close around this shaft and thus exclude the exterior air. The other end has a hole cut through it for the passage of the glass tube which connects the palladium asbestos tube of the hydrogen purifying apparatus to the hydrogen wash apparatus described by Loomis. Through the top, toward the back, there are three holes for the admission respectively, of the wires used in connecting the cells with the potentiometer, of the wiring for an electric lamp for heating the air when necessary to about 2 6 ' , and of the thermometer held by a stopper. T h e purpose o j the lamp rejerred to is to heat the a i r enclosed b j * the coiier to about 26' arid thus prezent the cogadettsatioTz, i~ the u p p e r parts o j the calomel electrodes, of the water Ltapor above the solutions oj potassium chloride and avoid changes in the concentrations and electromotive forces. The air-fan referred to serves t o circulate the air and prevent superheating above any one battery. An ordinary thermometer is considered sufficiently accurate for this air space, and no thermo-regulator is necessary as the lamp is not used when the room is warm enough to heat the
252
IT-. 1:. Clarke, C’.N. -lI.,t.is a n d 5. F..ici’ee
air of the bath sufficientlv to interfere with the regulation of the heating of the oil The lamp used is blackened with a heavy coating of lamp black and shellac so as to prevent radiation For the study of a calomel batter>-over a period of several years the form used by Myers and Acree is well adapted. It differs from the form used by Loomis in that the upper tubes of the single electrodes are constricted and sealed instead of being closed by means of ground-glass caps containing stopcocks. To study a battery for a short time and to empty and refill seemed highly desirable. To d o this with the f o r m used bjs -1Ijers aiid Acree i t wmiild be iieccssarg’ after two or t h e e jilliiigs to seal a iicw t o p to each electrode; as siiclz a process zncit i’s coizsidcrable yisk qt breakiiig aii e x p c i i s i x piece o j apparatzis, i t Lias tlzoitgltt bcht to i m d i j y the .torwz iiz such a way tlaat i t corild bc iiscd Icpcatedly witlzozit this daiigci’ atid at the same t i m e w i t h i i t pciwittiiig the lcaki$ig i i i o j otl as d i d tlze j o r w zised bj Lootiiis T17itlz titi., idea i i i miiid tlic j o i u z i H ) illustiatcd iiz J i g . I was J c a s c d . The essential differences consist in the use of ground-glass stoppers for closing each electrode and in using sealing-was to fasten the glass caps down upon the bulbs shown in Fig. I . This form of battery and that used by Myers and Acree contain much wider tubes connecting the single electrodes than Irere used by Loomis. This dcri’cascc tlzc clciti ita1 rcsistaiic-e aiid also p e i i i r i t s i t z o i t rapid di-ffusioii aizd iztralliillciif q r cqrrzlzbi izili?. Several modifications of the hydrogen apparatus used by Loomis are described by Alyers and Xcree. The use of p z i i c gziiiz stopper\, of the device for prex-enting the hydrogen electrodcs from drying while in storage during the summer, and of the shori-circuiting ”crow-foot” are important changes which they introduced The storage apparatus for the hydrogen ekctrcdes is illustrated in Fig z and Fig 3 An ordinary crjstallizing dish. Aij about hall filled nith o I -1HC1 is placed in a desiccator. F,which contains enough 0.1 HC1 t o rise to about of the height of the crystallizing dish. The tube D, passing through a stoppered h d e , j)in the side
A Study of the Hydrogew Electrode, E t c .
253
254
W . F . Clarke, C. N . ,Wyers and S. F . Acree
of the desiccator, and so bent as to have an outlet under the surface of the acid in the crystallizing dish, serves as an inlet for hydrogen gas, while the tube E serves for its escape. Fig. 3 shows the framework which supports the electrodes so that they are about half submerged in the acid in the crystallizing dish. B, B, and B2 are the three legs of a glass tripod, upon which, by means of the hook H, and the supporting gold wires 1’-I, 2 ’ - 2 , 3’-3, are hung the circular glass disks A’ and A”. These disks have cut into their edges indentations suitable for holding the electrodes, as illustrated by E ; after the set of electrodes has been placed in position in the disks a gold wire is bound around them in such a way that they will not slip if handled with ordinary care. The tripod is now placed in the desiccator so that the feet rest upon the shoulder B and B1in Fig. 2 . When the electrodes are not in use, and yet it is desired to keep them active, the wires C, C1, C2, CB, etc., are short-circuited in a cup of mercury outside the desiccator. Inside the desiccator each of these wires is introduced into a different capillary tube carrying a hydrogen electrode. When it is wished to intercompare the hydrogen electrodes, the wires are removed from the cup of mercury; the wire leading to any one electrode (which serves as a standard) is connected by means of a cup of mercury and wiring to the potentiometer; the other wires are similarly, one a t a time, connected to the potentiometer and the intercomparison of the potentials of the electrodes is made. Table VI shows that the deviation among a number of electrodes stored in this way for eleven months was generally less than 0.0001 volt. This is highly satisfactory as electrodes left dry even three months vary from 0.0001 volt to 0.06 volt, and cannot be used at all without replatinization. The barometer is a Fuess normal and carries verniers for correcting for capillarity. The connection between the hydrogen electrode and the calomel is made by means of the sand-filling device of Bjerrum as described later. With regard to the calomel electrode a little change in
A Study of the Hydrogen Electrode, Etc.
255
the preparation and introduction of the material has been made from the methods used by Loomis. I n the preparation of the calomel the precipitation of the mercurous chloride is effected at present by the addition of a dilute solution of potassium chloride to the solution of mercurous nitrate. The form of calomel battery has been modified as described. Some of this work is concerned with the study of batteries which were made up by Dr. Myers. Two of them designated as B-IV and B-V were made up in May, 1912,while B-I1 was made up May I, 1913. The former two have not been refilled during this time and have had a history worth considering a t length. Table I shows the data for battery B-IV0.1 KC1-HgC1-Hg, part of which has been presented by Myers and Acree. The notations “with wires,” “without wires,” etc., refer to the readings below; the notations “with wires” means that the crow-foot short-circuit device was in use. “Without wires” means it was not in use. The notation “readings made while storage cell was connected to charging switch” means that wires from the storage cell were connected to the switch leading to the charging current. Although the switch was open there was a leak sufficient to affect the galvanometer, as the readings show. “Changed short-circuit to wood cups” means that in place of the “crow-foot” there was introduced a set of wood cups containing mercury; these were connected by wires to the capillaries in the batteries. The cups of mercury were now short-circuited with one another by means of a frame of brass. “Changed short-circuit to glass cups” means that glass cups of mercury were used in place of wood, it being found that the wood cracked and let the mercury escape. On the whole the figures show that the best system is the “crow-foot” used by Myers and Acree. Table I1 shows the data battery B-V -0.I N KC1-HgC1Hg; some of the readings were made before this present study was begun. The notes and notations explain some of the variations in the readings.
1.t'. F . Clarke, c'. IV. M y e r s aMd S. F . Accree
256
TABLEI-BATTERY KO.IY,
0.1-1-KC1-HgCl-Hg Prepared May 2, 1912 The first values recorded were obtained 18 hours after preparation (Values expressed in millionths of a volt) Electrode I used as standard With wires
____ ~-
Date _ _ - ~ -
2
4
- .~
~
~
~-~-
3
j
G
7
8
-
-__
~~
9
1
0
~
~
1912
3 4 6
JIaY May May May May _____
7
9
--
140 11 I5
41 5 4
IO 12
38
50
25
5j
8 80 4 7
33
2
30 16
2 12 IO
15
90
45
26
8
6 1 1 13 2 I 1 1 I O __--_9_ 9 ____ Without wires 50 60 __ 15 -j o 60 With wires 3 j 7 12
7
IO
180 4 5
9
11
IO
3 0
80
75
~
~
May
11
May
22
I
I /
IO
9
I
I913
IO 1
Xpr. 11 NofE.-Readings made while storage cell mas connected to charging switch - - ~_ _ _ ~ _ _ ______ -~_ _ Nov. -p_-p14--2_44 2 _ 16 2_4-463636 46 Storage cell disconnected _N o v . 14 23 2 8 1 0 6 8 2 8 I 9 h-ov. 2 0 5--_9-_ 9--4 2 _' _1_ 4 _____--1 1 0 9 4 Changed short-circuit to wood cups ._ ___-___ 5 I O 11 o 3 Dec. 14 16 j g 23 2 0 2 Dec. 2 2 32 11 23 14 14 34 5 1 26 ~~
-
__p
-
-
I914
Feb. 9 Feb. 23 Mar. 2 AMar. 7 ____ - Mar.--2 2 ~
~
29 19
54 I
j 2 2 I1
jg 2
25 -2 2 1 0 2 4 1 2 9 - 9 6 8 j
73 34 13 2
73 16 8 6
IS
Wire crow-foot_ replaced 3 27 35 33 5 2 27 j0 43 23
61 16 8
61 26 14 j
j 20
_
18
-3
32
27
17 14
~ 2. 0
2
1
6 1 ~
-
2
29
16 19
Equilibrium had not been
1 Bath had been out of use for a long time. attained when these readings were made. 2 The lyire leading to capillary had broken off from 3-0. 5. 3 The wire leading into capillary of electrode S o . 8 was not amalgamated and did not make contact.
TABLE 11-BATTERY B-I-0.
I -\-KC1-HgC1-Hg Prepared Ma?- 2 , I 9 I 2 First readings recorded 18 hours after preparation \ITaluesexpressed in millionths of a volt) Electrode I used as standard n'ith wires
Date 1912
Ala?- 3 May 6 Ma?- 221
Oct.
I;
Oct. 2 2 h-ov. 4 '9'3
2
6
2
4
3
2
3
0
0
2
2
6
S
i
/
I
-/
1
6
4
0
2
0
I
I
I
1
2
4
97 8
82 20
87 4
81
56
26
20
21
22
64 3
21
SOT. I2
0
2
I7
2
34
13 3
Sor. 20 SO\-.2 j Dec. 14
j
I0
-
0
11 2
4j 38 3
-
8
2
3
3
2
5
s
0
4
2 3
4 3
0
2
Jan. 7 Jan. 14 Apr. I I llay j SOT. 42 h - 0 ~ .j
7
*
5
2 4 8
1 0 j
16 18
8 16 16
16
2
2
I0
2
0 2
8
2
3
22
4
0
2
0
2
0
/
4
IS 16 3
2
0
2 2
3 5 3
.. ?
8
2
1
6
6
-
'9'1
Feb. I I Feb. 20 Mar. 2 Mar. I; Alar. 2-7 May 8'
I
4'
2
6.3
>
58
-
3 33
~
4 3
3
0
3
78
6;
82
6 4 1
6;
4 68
-An examination of the 1-alues shown in TableL; I and I1 shows t h a t within four daj-s from the time of preparation of Battery had been used in making measurements, solution \\-as disturbed. Iieadings made while storage cell was connected to charging switch, short-circuiting arrangement \vas liacl. Better short-circuiting. 4 Batter>- had been used to make intercomparisons. Solutions were clistubed.
W . F . Clarke, C . N. Myers a?.zd S.F . Acree
258
the calomel batteries, the potentials of the electrodes will differ among themselves by not more than 2 0 millionths of a volt. The values found in the cases in which the crow-foot short-circuiting was maintained show that this system tends to bring about a closer agreement of the electrodes than is obtained when there is no short-circuiting. The close agreement between the electrodes at the end of two years shows the constancy of the system.
TABLEI11 (U)-COMPARISOKOF BATTERIESB-IV KCl-HgCl-Hg B-IV against Cell I B-V after 18 months
___________________
~~~
AND
B-I',
0.1 i"\i
B-IV
-
~~
~
XVithout wires
Same date without wires
I 2
0.00000j
0.000028
0.000001
o.00003 2
o.00003 I
0.
3 4
0.000029 0.00002 I
0.000017 o.000025
0.000000
5 6 7 8
0.000025
0.000012 0.000022 o.00003I
Cell
0.000012 0.000023
9 IO
o.000007 o.000004 0.000002
0.00000j
0.000014
0.000038 0.000010 o.000024
ooooog
o.00002 7
-
o.000007
,
o.000006 -
I
TABLEI11 COMPARISON OF BATTERIESB-IV AKD B-V, 0.1 N KCI-HgC1-Hg B-1- against Cell I of B-IV B-IV against Cell I of B-V _ - ~ _ _ ~___
_
~
_..__
- _
Cell
_
_ ~ _
1 I
I
o.000009
0 000001
2
0.000012 0.000000 0.00000j
0 00000~
3 4 5 6 7 8 9 IO
0 00000~
000002
o.000004
0 0 0 0 0
0.00000j
0 000000
0 . ooooog
0.00000 I
0.000008
0.000006
l
o
000000
000001 000000
000001
000003
A Study of the Hydrogen Electrode, Etc.
259
TABLE IV-BATTERY No. VI, 0.1 S KCl-HgC1-Hg (Values expressed in millionths of a volt) Electrode I used as standard With wooden cups of mercury - . -~
--
~
Feb. 11 I5 20 32 45 j j 20 2 2 I Feb. 13 25 2 6 ___ 38 I 43 10 __ 28 23 Unblackened 32 c. p. light placed in air space above bath Feb. 18 I j 19 35 33 31 5 11 Feb. 20 Feb. 2 3 2 j 28 1 54 51 19 14 8 Mar. 2 '4 - 4 2 52 34 I 8 I8 Blackened 8 c. . p. light in air space Mar. 8 5 32 28 32 29 I 3 3 12 4 Mar. 2 2 6 1 1 25 30 33 7 11 9 1 29 I Z 1 14 14 22 51 32 0 May I May 8 1 18 2 1 1 I j 5 g 35 9 6
37 54
~
i:
~
,
I3 2 20
31
260
VI;. F. Clarke, C . K.Xj'evs
a7id
S. F. Acree
Table 111 ( a ) and Table 111 ( b ) show the comparison of the electrodes of one battery against a single electrode of another. It should be noted that the effect of the use and non-use of the short-circuiting system is marked. The values shown in Table I11 ib) were obtained after the short-circuiting system had been in use for some time immmediately previous. Tables IV and V give the history of Batteries B-IV and B-VII. These batteries are the latest form, i. e . , they have ground-glass stoppers for each electrode and protecting glass covers. The values shown are not so good as those found for batteries B-IV and B-V, but this is due to better technique in preparing the materials for B-IV and B-V. So far as the form of battery is concerned, there seems no reason to believe the forms used for B-IV and B-V are either more or less preferable than those used for B-VII, so far as constancy and reproducibility of the electrodes are concerned. It is interesting to note the effect of the use of unblackened 32 . c. p. light and of blackened (to diminish radiation) 8 c. p. light for heating the air-space enclosed by the glass and wood top above the bath. Attention is directed to the fact that in the case of B-VI1 the crow-foot was used throughout its history, while for B-VI wooden cups of mercury were used. The results in the case of B-VI1 are rather better than for B-VI. This is believed to be due to the difference in the method of making the reading. In the case of B-VI the readings were made through the cups of mercury, each of which was connected, by means of a screw soldered to a wire, to an electrode. The soldering of the wire to the screw is believed to have brought about resistance differences which affected the readings on the potentiometer considerably. It will be noted that the crow-foot system was made use of in the case of B-VI previous to the last two sets of readings and that the values for the potentials of the different electrodes approached a more constant value. Table VI shows the history of a set of hydrogen electrodes. Some of the values shown were obtained previous to the time this present study was begun. TJllenever the
A S t u d y of the Hydrogen Electrode, E t c .
261
TABLE VI-Hz-Pt-o.r AYHCI. STOPPER I Electrodes were placed in 0.1 S HC1 after being left in box three months. H P allowed to run about five hours before first record was taken. Electrode I O chosen as standard ues in millionths of a volt) -
-~
-
-
p
-
Date 9 8 _ _ _ _ _ ~
5
6
j -~
~
~_
--
p
4
3
p
2
I
-~
-~
1912
Jan. 18 186 820 470 700 62930 9 j 6 588x0 155 Jan. 19 232 178 168 169 4988 1736 3445 696 Jan. 2 0 173 - 8 1 ~ 5 0 1195 1720 405 ~ ._2654--895_
1
2558 820 525
~Began use of “Crow-foot” replatinized and . . electrolyzed -__ Jan. 24 I j 35 40 98 31 53 15 15 Jan. 26 6 IO IO IO 46 9 5 25 Jan. 2 7 7 3 o j IO 6 0 40
J ~ c 3.-1
- 1 0-_
6
6
_~
-~
~~~~
3
28
6
-
_
6
-
p_
47 I3 2
3 7 ‘~4 -
-
~
Rubber dissolving in acid as shown below Feb. I 42 2 j 46 14 46 58 44 7 8 44 Apparatus cleaned. Electrodes and stoppers boiled in distilled water Feb: 1 1 1 I j 12 6 IO I8 I2 2 j 1 25 Feb. 6 4 4 I 3 3 8 2 2 5 Stove in bath burned out and temperature drops 6 degrees Feb. 9 6 3 3 j 4 2 6 6 I 1Vithout wires ____. Feb. 17 25 IO I j I j 14 12 9 I j -I Feb. 19 20 3 Aj- 15 26 26 25 34 27_ Rubber dissolving in acid Mar. 6 jo-60 130 jj I I ~ 90 5 0 140-l 56 Electrodes boiled in distilled water and placed in constant pressure desiccator _
-
-
_
p -
_
_
-
_
1913
-~
I
-:
June I 20 3 1 j 2 3 I Apr. 2 j j 3 6 6 ~- . 7 54I 3~ 1 Electrodes stored in desiccator until March 29, 1914, then placed in 0.1 HCl and exposed to stream of Hz -~
~
~-
~
~
~
-
p
-
p
-
‘914
_ _ 31 - 24 Mar. ~ ~2 0 _ 33 _ 192 105 46 44 70 Electrodes replatinized. Placed in bath May 6 ___~____ _ &lay 8 g 20 12 6 8 25 9 12 ~
~
p
~
~
__ _
11
~p
~
25
This electrode was broken and repaired, giving result shown on next date.
262
W . F . Clarke, C. N . Myers and S . F . Acree TABLE VII-Hg-HgC1-o. (Bar. in Atms.,
1.010;
~
I -4’KC1-o. I N HC1-Pt-H2 Cor. in- e.- rn. f., -0.00023) - -~ -~
~
427341 0 427342 0.427336 0 427342 0 427335 0 427342 0 427340 0 427335 0 427338 0 427340
~~~
~~
_____
0
Other similar averages o 427106 (a) 0 427109 ( b ) 0 427109 (c) 0.427108 (d)
Av- 0 427339 Bar. Cor. o 00023 AV.
0
Av.
427109
Cor. av. o 426919
o 427108
Cor. av. o 426918
electrodes were replatinized they were boiled with distilled water for several hours after the replatinization. Table VI1 shows a comparison of a fresh calomel battery and a hydrogen battery without use of a device for eliminating contact potential. These values were obtained by Myers and are reported here for the sake of comparison with a system in which there was used for elimination of contact potential a saturated KC1 solution. The values ( a ) , ( b ) , (c), ( d ) are averages of sets of values obtained at other times. All readings shown were made before the Fuess normal barometer was obtained, and the “ Cor. Av.” notation means that the readings of the old barometer were not correct, hence the values noted “Av.” need the correction applied. TABLE VI11 0.1h-KC1-HgC1-Hg sat. KC1 0.1i‘i HCl-Pt-H2 B-V-Bar. in atms., o.ggeBar. Cor. = 0.00026 0,39971 0.39972 0.39980 0.39974 0,39972
0.39981 0.39976 0,39971 0.39972 0.39973
Av. 0,39974 Bar. cor. 0.00026
, ~
o.40000
Cor. av. 0.39982
A S t u d y of the Hydrogen Electrode, Etc.
263
Table VI11 shows the values obtained when saturated KCl is used to eliminate contact potential. The values obtained agree closely with those of Bjerrum. The results for the system 0.1N HC1-HgC1-Hg were about the same as those obtained by Myers and Acree.
Summary I . The object of this investigation has been to devise better methods and apparatus for the accurate measurement of the electromotive force of hydrogen and calomel electrodes. Chief among the improT7ements in apparatus were ( I ) a new heating and stirring device for the oil bath which permits a more accurate regulation of temperature and electromotive force; ( 2 ) an air bath above the oil bath to heat the air and prevent evaporation from the liquids and changes in concentration and electromotive force; (3) a new form of calomel battery which can be filled, emptied and cleaned very quickly and easily without danger of breakage and which gives as accurate values as those used by Myers; (4) a storage reservoir in which hydrogen electrodes can be kept active in hydrochloric acid in an atmosphere of hydrogen and kept constant in electromotive force to within 0.0001 volt. The preparation of very pure calomel by the use of potassium chloride and the use of weight methods in lieu of volume methods for the preparation of the hydrochloric acid have increased the ease of making satisfactory solutions. 2 . By the use of these improvements in the apparatus and methods i t is possible to obtain much more easily than formerly a reproducibility of 0.01millivolt with the hydrogen and calomel electrodes, 3. The 0.1N KC1-HgC1-Hg system seems to be constant and reproducible to within 0.01millivolt. In addition it is a comparatively stable system. 4. The hydrogen electrode has proven to be an easily reproducible and constant system, extremely sensitive to small amounts of impurity, thus indicating its rapidity and accuracy for determining small hydrogen ion concentrations,
264
W . F . Clarke, C. N . Myers and S. F . Acree
which would render it very valuable in determining the hydrogen ion concentration in cases of catalysis by acids. j. It is found that the system Hg-HgC1-o. I N KC10 .I N HCl-Pt-H2 is an extremely accurate and constant element. It undoubtedly gives one of the best intercomparisons that can be obtained. 6. The 0 . I N HCI-HgC1-Hg system is found to be rather unsatisfactory over long periods of time. The intercomparison of the individual electrodes indicate that its constancy continues over only a short period of time, due to the formation of HgC12 and Hg, with a possible further decomposition. The values obtained during the past year support the evidence previously found in this laboratory. 7. An accuracy of 0.01millivolt was obtained when the following precautions were observed : a. The materials were most carefully purified. The mercury was purified by washing about five hundred times with three percent nitric acid. It was rinsed with water and let stand under concentrated sulphuric acid. It was distilled three times in a current of air. This was used for making calomel and in the calomel electrode. The calomel was made by dissolving mercury, purified as indicated, in N 5 nitric acid and using an excess of mercury. To the solution of mercurous nitrate a solution of thrice recrystallized potassium chloride was added. The calomel was shaken out with water repeatedly and filtered and washed to get rid of nitric acid and potassium nitrate. It was then shaken with approximately N I O potassium chloride to get rid of the excess of water. Finally it was Saturated with carefully standardized N / I O potassium chloride prepared from the ignited recrystallized potassium chloride. The N I I O potassium chloride was prepared from potassium chloride recrystallized three times and ignited to constant weight. Conductivity water was used in making up the solution. The N / I O hydrochloric acid was made from redistilled hydrochloric acid. The hydrogen was prepared by elec-
.-i Stud31 o j the Hydroge?z Electrode, Etc.
2 65
trolysis of dilute sodium hydroxide, using nickel electrodes. The hydrogen was passed through a palladium asbestos tube and a washing apparatus containing N I O hydrochloric acid. b. A temperature relation in the oil to within 0.01' was maintained for a t least two days previous to making readings. The temperature above the bath was roughly kept a t 26'. It was intended simply to prevent the condensation of liquid in the upper part of the apparatus which protruded above the oil. (2. When not in use for several weeks the hydrogen electrodes were stored in the container illustrated in Figs. 2 and 3 . They were kept saturated with N / I O hydrochloric acid and hydrogen. d . The hydrogen was bubbled through the N I I Ohydrochloric acid against the electrodes. It was believed that this is more efficient than simply maintaining an atmosphere of hydrogen above the hydrochloric acid. e. The individual electrodes of the calomel (and hydrogen) batteries were short-circuited together by means of the wire crow-foot. This produced and maintained a very close agreement. Johns Hopkins University Baltimore
